Key points

Introduction

Ethnicity is a biological and a social construct, which includes ancestral genes and cultural, geographic, and socioeconomic characteristics shared by populations. In contrast, race refers to genetically homogeneous groups of people, which is hardly the case for the human race. The US population is heterogeneous even within its ethnic groups, as shown when ancestry informative markers (AIMs, genetic markers differentially expressed in founding populations) were examined in the LUMINA (Lupus in Minorities: Nature vs Nurture) study; in Texan Hispanic patients (mostly of Mexican origin) nearly half of their AIMs were Amerindian (48%); in contrast, the Puerto Rican Hispanics had a larger proportion of white/European (35%) and African AIMs (45%); whites and African Americans had about 20% of AIMs of the opposite group. The term Hispanic (in the United States) applies to all individuals whose origin can be traced to a Spanish-speaking country, and it is a very heterogeneous population group. On the other hand, approximately 40% of Latin Americans are mestizos, or of mixed European and Amerindian ancestry, but there is great variability within the individual countries. The remaining Latin Americans are of European, African, or Asian ancestry; despite this heterogeneity, in the United States, these individuals may be grouped with the Hispanics, although some distinctions may be made (such as Hispanic blacks). It has been shown that there is a tight association between race/ethnicity (race is used in this review if used in the publication being cited) and socioeconomic features, with minorities having a lower socioeconomic status (SES) than the white majority. Thus, differences in disease expression, intermediate-term (disease activity), mediate-term (damage accrual, work disability, health-related quality of life), and long-term (mortality) outcomes relate to both components of ethnicity, and it is difficult to disentangle the influence of each one.

In the following sections, the influence of race and ethnicity on disease expression, disease outcomes, response to therapy, and renal replacement therapy is discussed.

Introduction

Ethnicity is a biological and a social construct, which includes ancestral genes and cultural, geographic, and socioeconomic characteristics shared by populations. In contrast, race refers to genetically homogeneous groups of people, which is hardly the case for the human race. The US population is heterogeneous even within its ethnic groups, as shown when ancestry informative markers (AIMs, genetic markers differentially expressed in founding populations) were examined in the LUMINA (Lupus in Minorities: Nature vs Nurture) study; in Texan Hispanic patients (mostly of Mexican origin) nearly half of their AIMs were Amerindian (48%); in contrast, the Puerto Rican Hispanics had a larger proportion of white/European (35%) and African AIMs (45%); whites and African Americans had about 20% of AIMs of the opposite group. The term Hispanic (in the United States) applies to all individuals whose origin can be traced to a Spanish-speaking country, and it is a very heterogeneous population group. On the other hand, approximately 40% of Latin Americans are mestizos, or of mixed European and Amerindian ancestry, but there is great variability within the individual countries. The remaining Latin Americans are of European, African, or Asian ancestry; despite this heterogeneity, in the United States, these individuals may be grouped with the Hispanics, although some distinctions may be made (such as Hispanic blacks). It has been shown that there is a tight association between race/ethnicity (race is used in this review if used in the publication being cited) and socioeconomic features, with minorities having a lower socioeconomic status (SES) than the white majority. Thus, differences in disease expression, intermediate-term (disease activity), mediate-term (damage accrual, work disability, health-related quality of life), and long-term (mortality) outcomes relate to both components of ethnicity, and it is difficult to disentangle the influence of each one.

In the following sections, the influence of race and ethnicity on disease expression, disease outcomes, response to therapy, and renal replacement therapy is discussed.

Disease expression

Differences in disease expression, including disease activity and damage accrual, as a function of ethnicity have been well documented ( Table 1 ). In general, patients from ethnic minorities not only develop systemic lupus erythematosus (SLE) more frequently, but tend to have an acute disease onset, present a greater number of, and more severe, clinical manifestations, show higher disease activity, have a higher risk of relapses, accrue more damage (and at a faster pace), and show higher mortality than whites. SLE also tends to present at a younger age in nonwhites (Hispanics, African descendants, and Asians).

Distinct clinical SLE patterns are found among patients from different ethnic groups. Whites, for example, show cutaneous manifestations more frequently, particularly, photosensitivity and malar rash. In contrast, African descendants experience discoid lupus more frequently, except for the Gullah population of the Sea Islands of South Carolina, African descendants with minimal genetic admixture, and in whom most integument manifestations are described. Hispanics, African descendants, and Asians show renal, hematologic, serosal, neuropsychiatric, and immunologic manifestations more frequently than whites. Lupus nephritis (LN), the most worrisome SLE manifestation, is significantly more prevalent in mestizos, Hispanics, Africans, and Asians. For example, in the LUMINA cohort, LN occurred more frequently and tended to occur earlier in the African Americans (62%) and Texan Hispanics (62%), whereas in the Puerto Rican Hispanics, LN occurred at a rate comparable with the whites (26%). When the contribution of admixture, preferentially reflecting the genetic component of ethnicity, and SES to renal involvement was examined in these patients, admixture explained a larger proportion (African primary but also Amerindian) of the ethnicity-dependent variance than SES (36.8% vs 14.5%), whereas both accounted for an additional 12.2% of the variance. Sánchez and colleagues showed in a larger study that Amerindian ancestry is a predisposing factor for LN occurrence, whereas Richman and colleagues reported that European ancestral genes are protective.

In the GLADEL (for Grupo Latinoamericano de Estudio del Lupus or Latin American Group for the Study of Lupus) cohort, the African Latin American patients and the mestizo patients tended to develop renal disease more frequently and earlier than the whites, and this was statistically significant only for the mestizo patients.

Contreras and colleagues in a study conducted in South Florida in which 213 biopsy-proven cases of LN in patients from 3 different ethnic groups (whites, Hispanics, and African Americans) were included, found less favorable outcomes in the African Americans and Hispanics. Renal events (doubling serum creatinine level or end-stage renal disease [ESRD]) or death occurred with a frequency 3 times higher in African Americans and 2 times higher in Hispanics when compared with whites. The Hispanics included in this study were predominantly Cuban and South American.

We have observed differences in disease expression in the 2 Hispanic subpopulations we have studied. The Texan Hispanics tend to have a more severe disease than the Puerto Rican Hispanics, as shown by higher levels of disease activity, more damage accrual, and more frequent occurrence of LN, psychosis, serositis, and thrombocytopenia. In turn, the Puerto Rican Hispanics tend to have integument manifestations such as photosensitivity, malar rash, and discoid lupus more frequently. The Texan Hispanics have a disease as severe as that of African Americans. The US Hispanics, particularly those of Amerindian ancestry, tend to also develop obesity, diabetes, hypertension, and the metabolic syndrome more frequently than do the non-Hispanic whites, which has also been proved to be the case in Hispanics from Mexico, as described in cross-sectional analyses of the SLICC (Systemic Lupus International Collaborating Clinics) inception cohort. In this study, Korean patients also showed the metabolic syndrome frequently although central obesity occurred less commonly than in the Hispanics.

Asian patients with lupus also experience a more severe disease than whites, with higher rates of neuropsychiatric involvement, but comparable with those reported in Afro-Caribbean patients. Asian patients with lupus also develop LN at rates comparable with that observed among Afro-Caribbean, African American, and Texan Hispanic patients (58%–69%). Compared with whites, the risk of renal disease in Asian patients is, overall, 3 times higher, but there is significant variability among various Asian populations; higher rates are observed among Chinese (57%), Indians in Malaysia (70%), and Malaysians (70%–76%), but not in Pakistanis (33%). Indian patients in Malaysia also experience neuropsychiatric involvement frequently (36%–39%). In agreement with these findings, Asian patients with lupus in Canada and the United Kingdom experience significantly more renal involvement than whites ; however, like in Asia, the prevalence rates are not uniform, with higher prevalence observed in Chinese patients (100.3 per 100,000 population) than in patients from the Indian subcontinent (21.4 per 100,000 population).

Disease activity

Disease activity either at disease onset or over the duration of the disease, regardless of how it is measured, has been found to be higher among nonwhites (African Americans, Afro-Caribbeans, African Latin Americans, Texan Hispanics in the United States, mestizos in Latin America, and Asians in Australia), although the differences have not always been statistically significant. At disease onset, there seems to be an important genetic contribution to disease activity, but over the course of the disease, socioeconomic features seem to exert a more important role, as shown in the LUMINA study.

As in the LUMINA study, in the SLICC cohort, whites had significantly lower disease activity at each yearly assessment over a 5-year follow-up period compared with nonwhites, with the difference in disease activity mainly driven by Asians and Hispanics. Golder and colleagues, in a single-center study conducted in Australia, found that Asian patients have a disease significantly more severe with higher time-adjusted mean Systemic Lupus Erythematosus Disease Activity Index (SLEDAI) scores, higher maximum SLEDAI score, and more frequently persistent disease activity compared with non-Asians (most of them whites). Furthermore, Asians were more likely to have persistently active disease, as defined when patients had at least 2 consecutive visits with an SLEDAI score of 6 or greater. Similar data have not emerged from Asian countries.

Damage accrual

Damage or the irreversible consequences of the disease or of the therapies used to abate its manifestations has been extensively studied. The SLICC Damage Index (SDI) documents irreversible organ system manifestations present for at least 6 months from disease onset. Factors associated with damage occurrence include older age, male gender, African American and Texan Hispanic ethnicities, poverty, disease duration, disease activity, number of American College of Rheumatology criteria met, glucocorticoid use, and abnormal illness-related behaviors. In addition, Texan Hispanics accrue damage more rapidly than patients from the other ethnic groups. When these patients were compared with patients from Northern Spain (Spaniards), SDI scores were comparable at baseline, but at year 4, their scores were higher, despite having shorter disease duration at enrollment than the Spaniards. Previous damage has been found to be a risk factor for the occurrence of further damage. It seems that the combination of a more severe disease and poor SES acts synergistically, contributing to greater damage accrual among minorities in the United States and elsewhere.

Differences in the pattern of organ damage have also been observed. For example, neuropsychiatric damage occurs more frequently among whites, as described in the Maryland and LUMINA cohorts. Renal damage occurs more frequently among African Americans, Afro-Caribbeans, and Texan Hispanics, as documented in several studies ; however, in the CLU (Carolina Lupus) study, the association of renal damage and nonwhite ethnicity was no longer significant after adjusting for age and household income, further reinforcing the role of SES in damage accrual in lupus. Furthermore, Texan Hispanics and African Americans tend to develop ESRD more frequently than patients from other ethnic groups. In African Americans, genetic factors have been shown to contribute to the risk of LN-ESRD; the apolipoprotein L1 gene ( APOL1 ) nephropathy risk alleles G1/G2, common in African Americans and rare in European Americans, are strongly associated with the risk of LN-ESRD as well as the time to progression to ESRD in African Americans. African Americans also tend to have a less favorable outcome when they undergo renal engraftment, which relates to a higher number of risk factors for allograft failure (higher rate of cadaver donors with greater degree of HLA mismatch and higher levels of reactive antibodies against HLA compared with Hispanics and whites). Moreover, lower income levels have been associated with higher risk for graft loss and death among transplanted African Americans.

Integument damage occurs more commonly among African Americans, ocular damage among Texan Hispanics, and malignancy among whites. Diabetes and alopecia were significantly more frequently among the CLU African Americans ; damage in the musculoskeletal system, particularly avascular necrosis, was commonly found in African Americans from the Hopkins cohort but not in other published studies.

Studies performed in different Asian subgroups (Chinese, Koreans, and Arabs) in their country of origin have shown that these patients accrue damage at similar rates to whites, but they tend to accrue more renal but less cardiovascular and ocular damage. In addition, there is variability in the patterns of damage across different Asian countries, with Chinese patients having a high prevalence of premature gonadal failure, whereas patients residing in southern and western Asian countries develop more integument damage, which probably relates to an increased exposure to ultraviolet B rays.

As for the Asians residing in western countries, overall damage has been generally reported to be higher than that observed among the respective white populations, but this has not been observed in Canada. Asian patients in Canada had the lowest damage score, and at an average of 10 years of disease, 59% had not accrued any damage compared with 35% of Afro-Caribbean and 42% of white patients. However, Asians in Canada had a similar SES to that of whites, and it is possible that this may have protected them.

Work disability

Illness-related work disability has profound repercussions at the individual and societal levels, because work loss leads to a diminished self-esteem, income loss, inability to accumulate assets for retirement, and social isolation. Three to 15 years from diagnosis, 15% to 51% of patients with SLE report ceasing employment. This outcome has been associated with several different demographic, disease-related, and job-related factors, such as older age, lower levels of education, African American ethnicity, poverty, longer disease duration, higher disease activity and damage scores, fatigue, fibromyalgia, anxiety, neurocognitive involvement, thrombotic manifestations, and higher physical job demands. Thrombotic events have been associated with an increased risk of acute work loss, whereas musculoskeletal manifestations and neuropsychiatric events have been associated with delayed work loss.

When the risk factors for self-reported work disability in patients from the LUMINA cohort were examined, African American and Hispanic (from Texas) ethnicities were associated with the occurrence of work disability over the course of the disease in univariable analysis; however, poverty along with older age, male gender, disease activity, damage accrual, and disease duration, but not ethnicity, were retained in the multivariable models, supporting the importance of socioeconomic factors on the development of this intermediate outcome and the association between minority status and poor SES in the United States and probably elsewhere. Similar findings were obtained by Utset and colleagues, and like in LUMINA, African American ethnicity was not retained in the multivariable analyses, except when neurocognitive impairment and damage were removed from the model; then, a trend was found.

Work disability has rarely been studied in Asian patients with lupus. In a cross-sectional questionnaire study performed in Chinese patients, a high cumulative incidence of work disability was found (36% at 5 years of diagnosis); in this particular study, older age, fatigue, and more active disease were independent predictors of work disability. In another cross-sectional Chinese study, the prevalence of SLE-related complete work disability after median disease duration of 9 years from onset was 16%. SES (lower levels of education) and disease factors (longer disease duration and history of having pleurisy) were independently associated with increased risk of work disability. Neuropsychiatric involvement and a lower utility score (derived from the Euroqol 5 dimensions, a standardized measure of health status) were not significantly associated with work disability in this study.